A detailed scalable continuous flow protocol is presented to synthesize an aryl fluoride from an aryl amine through the Balz-Schiemann reaction.
The demand for aromatic fluorides is steadily increasing in the pharmaceutical and fine chemical industries. The Balz-Schiemann reaction is a straightforward strategy for preparing aryl fluorides from aryl amines, via the preparation and conversion of diazonium tetrafluoroborate intermediates. However, significant safety risks exist in handling the aryl diazonium salts when scaling up. In order to minimize the hazard, we present a continuous flow protocol that has been successfully performed at a kilogram scale that eliminates the isolation of aryl diazonium salts while facilitating efficient fluorination. The diazotization process was performed at 10 °C with a residence time of 10 min, followed by a fluorination process at 60 °C with a residence time of 5.4 s with about 70% yield. The reaction time has been dramatically reduced by introducing this multi-step continuous flow system.
The Balz−Schiemann reaction is a classic method for replacing the diazonium group with fluorine by heating ArN2+BF4− without a solvent1,2. The reaction can be applied to a wide variety of aryl amine substrates, making it a generally applicable approach to synthesize aryl amines, which are frequently utilized for advanced intermediates in pharmaceutical or fine chemical industries2,3. Unfortunately, harsh reaction conditions are often employed in the Balz-Schiemann reaction, and the reaction generates potentially explosive aryldiazonium salts4,5,6,7,8. Other challenges associated with the Balz-Schiemann reaction are the formation of side products during the thermal decomposition process and its modest yield. In order to minimize the side product formation, thermal dediazotization can be performed in nonpolar solvents or using neat diazonium salts9,10, which means the aryldizanium salts should be isolated. However, the diazotization of aromatic amines is generally exothermic and fast, which is a risk associated with the isolation of the explosive diazonium salt, especially in large-scale production.
In recent years, continuous flow synthesis technologies have helped to overcome the safety issues associated with the Balz-Schiemann reactions11,12. Although there are some examples of diazotization of aromatic amines using continuous microreactors for deamination at positions para to aryl-chlorides, 5-azodyes, and chlorosulfonylation, these contributions were only reported on a laboratory scale13,14,15,16,17. Yu and co-workers developed a continuous kilo-scale process for the synthesis of aryl fluorides18. They have shown that the improved heat and mass transfer of a flow system would benefit both the diazotization process and the fluorination process. However, they used two separate continuous flow reactors; therefore, the diazotization and thermal decomposition processes were investigated separately. A further contribution was published by Buchwald and co-workers19, where they presented a hypothesis that if the product formation was proceeding through the SN2Ar or SN1 mechanism, then the yield may be improved by increasing the concentration of the fluoride source. They developed a flow-to-continuous stirred tank reactor (CSTR) hybrid process in which the diazonium salts were generated and consumed in a continuous and controlled manner. However, the heat and mass transfer efficiency of a CSTR is not good enough as a tube flow reactor, and a large CSTR cannot be expected to be used with explosive diazonium salts in large-scale production. Subsequently, Naber and co-workers developed a fully continuous flow process to synthesize 2-fluoroadenine from 2,6-diaminopurine20. They found that the exothermic Balz-Schiemann reaction was easier to control in a continuous flow manner and that the tubing dimensions of the flow reactor would influence the heat transfer and temperature control aspects – a tube reactor with large dimensions shows a positive improvement. However, the tube reactor's scaled-up effect will be notable, and the poor solubility of the polar aryl diazonium salt in organic solvents is troublesome for static tube reactors, which face a blockage risk. Even though remarkable progress has been established, there are still some problems associated with large-scale Balz-Schiemann reactions. Thus, the development of an improved protocol that would provide rapid and scalable access to aryl-fluorides is still significant.
The challenges associated with large-scale Balz−Schiemann reaction processing include the following:(i)the thermal instability of an accumulated diazonium intermediate over a short time period21; (ii) the long processing times; and (iii) the non-uniform heating or the presence of water in the diazonium fluoroborate, leading to uncontrollable thermal decomposition and increased by-product formation22,23. Additionally (iv) in some flow processing modes, an isolation of the diazonium intermediate is still required due to its low solubility14, which is then fed into a uncontrolled rate decomposition reaction. The risk of handling a large quantity of in-line diazonium salt cannot be avoided. Thus, there is significant benefit in developing a continuous flow strategy to solve the abovementioned problems and avoid both the accumulation and the isolation of the unstable diazonium species.
In order to establish an inherently safer production of chemicals in pharmaceuticals, our group has focused on multi-step continuous flow technology. In this work, we apply this technology to the Balz−Schiemann synthesis on a kilogram scale in a way that eliminates the isolation of aryl diazonium salts, while facilitating efficient fluorination.
CAUTION: Carefully check the properties and toxicities of the chemicals described here for the appropriate chemical handling of the relevant material as per the material safety data sheets (MSDS). Some of the chemicals used are detrimental to health, and special care must be taken. Avoid inhalation and contact with skin of these materials. Please wear the proper PPE during the whole process.
1. Preparation of feeds for continuous flow protocol
2. Continuous flow equipment setup
3. Continuous flow reaction processing
4. Distillation of organic solvents
The model reaction is shown in Figure 2. 2-Methylpyridin-3-amine (compound 1 in Figure 2) was chosen as the starting material to prepare 2-methylpyridin-3-fluoride (compound 3 in Figure 2) via the Balz-Schiemann reaction. The experimental parameters were systematically investigated by varying reaction temperature and residence time. Feed A is 0.35 M 2-methylpyridin-3-amine in THF. Feed B is pure BF3·Et2O at an 8.1 M concentration. Feed A and Feed B were first introduced into the premixing and cooling microflow module. Then, Feed C was charged into the dynamically mixed flow reactor to merge with premixed compound 1 and BF3·Et2O for a generation of diazonium salts. The solid was sufficiently dispersed to a slurry state in the dynamic tube reactor without tube blocking. The molar ratio of compound 1:BF3·Et2O: tert-butyl nitrite is 1:3.4:1.2. Finally, the diazonium slurry was fed into the thermal decomposition microflow reactor. Feed D was charged to the thermal decomposition zone at the same time as the diazonium slurry. The overall mixture was gathered into the product collecting vessel. Further purification was conducted in a batch manner.
Compared with the batch results (shown in Table 3), the reaction temperature of the flow experiments was successfully controlled at 10 °C while obtaining the crude product with >70% HPLC purity (Table 1, Entry 5), which was higher than what was obtained in batch processing (Table 3 and Supplementary File 1). The major impurities generated in the reaction are the hydrolyzed product/aryl hydroxide called imp-1 and the reduced material called imp-2 (Supplementary File 1 and Supplementary File 2). It is common to form imp-1 from moisture in the reaction system. Therefore, the specification of the water content of the solvent and the co-solvent was set below 0.5%. A typical experimental process batch process was the reaction of diazonium formation carried out in dry THF. To a stirred solution of 1 (10 g, 89.0 mmol) in THF (150 mL) at -10 °C under argon, BF3·Et2O (38.18 g, 266 mmol) was added dropwise. After 0.5 h, t-BuNO2 (10.17 g, 98 mmol) was added dropwise while maintaining an internal temperature of -20 to -10 °C. Solid precipitates were formed after 1 h. The solid was slowly transferred to 150 mL of n-heptane at 60 °C (safety note: Avoid isolation of the solid; experiments on a small scale revealed that it is unstable at 0 °C to room temperature). The residue was partitioned between organic solvents and HCl (1 M aq., 50 mL). The organic layer was washed with water, dried over MgSO4, and concentrated. Nearly 6.1 g of reddish-brown oily liquid was obtained with a 60% yield, analyzed by MS, 1H NMR, and 9F NMR (Supplementary File 3, Supplementary File 4, and Supplementary File 5).
Nearly 98% of conversion was reached within the 10 min residence time when the flow rate is 50 mL/min (sum of feed A to C; Table 2, Entry 2). Increasing the flow rate from 50 mL/min to 100 mL/min will lead to a lot of starting material remaining in the reaction (Table 2, Entry 3). Lowering the flow rate can lead to full consumption of the starting material (Table 2, Entry 1), but the production efficiency will be restricted. Therefore, a flow rate of 50 mL/min is chosen for scale up production with a daily capacity of 72 L/day.
Figure 1: The continuous flow processing flowchart. There are three zones: premixing, diazonium formation, and thermal decomposition zone. The internal volume of each reactor is 9 mL, 500 mL, and 9 mL, respectively. Please click here to view a larger version of this figure.
Figure 2: The synthesis of 2-methylpyridin-3-fluoride from 2-metyhlpyridin-3-amine via the Balz-Schiemann reaction. There are two major impurities generated in the process, one is the byproduct of hydrolysis imp-1 and the other is a reduced byproduct imp-2. Please click here to view a larger version of this figure.
Entry | Diazotization zone | Thermal decomposition zone | Conversion of 1 | HPLC purity of 3 | ||
Flow rate (mL/min) | Residence time (s) | Flow rate (mL/min) | Residence time (s) | |||
1 | 33 | 909 | 66 | 8.20 | 100.00% | 69.35% |
2 | 50 | 600 | 100 | 5.40 | 98.13% | 70.63% |
3 | 100 | 300 | 200 | 2.70 | 56.30% | 45.64% |
Table 1: The temperature effect of the diazonium formation process. The standard flow rates of pumps A, B, C, and D are 23.8 mL/min, 3.4 mL/min, 22.8 mL/min, and 50.0 mL/min, respectively. The residence time are listed as follows: a premixing zone of 20 s, a diazonium formation zone of 600 s, and a thermal decomposition zone of 5.4 s. The reaction temperature of thermal decomposition is 60 °C.
Process manner | Diazonium formtion temperature (°C) | Batch size | Conversion of 1 | HPLC Purity of 3 |
Batch process | -20 | 10 g | 100.00% | 66.62% |
Flow process | 10 | 0.5 kg | 98.00% | 70.63% |
Table 2: The residence time effect of the diazotization and thermal decomposition processes. The flow rate at the diazotization zone is the sum of Feeds A to C. The standard reaction temperature is 0 °C. The internal volume of the diazotization zone is 500 mL, and of the thermal decomposition zone is 9 mL.
Entry | Inner Temperature (°C) | HPLC Purity (%) | ||||
1 | 3 | Imp-1 | Imp-2 | Unknown Imp-3 | ||
1 | -20 | 1.84 | 60.80 | 9.98 | 0.00 | 1.53 |
2 | -10 | 0.53 | 58.56 | 6.17 | 0.00 | 2.08 |
3 | 0 | 0.00 | 70.99 | 9.16 | 1.12 | 1.31 |
4 | 5 | 0.06 | 64.85 | 0.00 | 3.54 | 6.74 |
5 | 10 | 0.00 | 70.63 | 0.00 | 0.00 | 5.84 |
6 | 20 | 0.00 | 56.07 | 0.00 | 0.00 | 5.22 |
Table 3: Comparison between batch and flow processes. The IPC results are shown in Supplementary File 1 and Supplementary File 2. Abbreviations: IPC = In-process control, HPLC = High performance liquid chromatography, Imp = Impurity.
Supplementary File 1: The HPLC IPC results of fluorination in the batch process. Please click here to download this File.
Supplementary File 2: The HPLC IPC results of fluorination in the flow process. Please click here to download this File.
Supplementary File 3: The MS spectrum of 3. MS (ESI), m/z calculated – 112.05 (M+H)+ and detected – 112.07. Please click here to download this File.
Supplementary File 4: The 1H NMR spectrum of 3. 400 MHz in CDCl3: δ8.32 (dt, J = 4.8, 1.5 Hz, 1H), 7.34 (ddd, J = 9.5, 8.2, 1.4 Hz, 1H), 7.20-7.09 (m, 1H), 2.55 (d, J = 3.0 Hz, 3H). Please click here to download this File.
Supplementary File 5: The 19F NMR of spectrum of 3. 376.5 MHz in CDCl3: δ-124.10. Please click here to download this File.
A continuous flow protocol of the Balz-Schiemann reaction has been successfully performed through a combination of a micro-channel flow reactor and a dynamically mixed flow reactor. This strategy features several advantages compared with the batch process: (i) it is safer with controlled diazonium salt formation; (ii) it is more amenable to a higher reaction temperature, 10 °C versus -20 °C; and (iii) it is more efficient without isolation of the diazonium intermediate, two steps in one continuous process. Specifically, the continuous flow equipment setup process is critical for this protocol, because besides the chemistry transformation, the reliability of a flow process is mainly dependent on the stability of the equipment. In brief, this protocol provides a technique for use with a wide range of aryl amine substrates, which are not feasible to diazotize and further react under batch conditions.
There are some experiences to be shared about flow technology development after this scalable Balz-Schiemann continuous flow project. When considering how to develop flow processes, the solid formation is an important parameter. In this situation, the following strategies could be tried. Firstly, (i) modify the protocol in batch-mode to maintain a homogenous phase throughout the reaction via changing reagents, solvents, or the temperature (ii) or choose a suitable flow reactor that allows for processing of slurries such as a dynamically mixed flow reactor. Another situation is that the feed of raw materials is a slurry. Normally, it is difficult to pump fine suspensions consistently and effectively in a scaled-up flow production. Pay attention to the pump’s ability to pump viscous liquid while handling a pressure drop. In addition, accumulation of particles in the reactor, especially in the microflow reactor, can lead to blockage, and ultimately failure. Therefore, combinations of flow reactor types with complementary advantages need to be considered to match the chemical attributes when carrying out continuous flow processing. Indeed, that is the driver, in this example, for the choice of two reactor types to realize this continuous flow process, which can be called the hybridized flow process.
There are still some limitations of this protocol. The heat and mass transfer efficiency of a dynamically mixed flow reactor is not as good as the microflow reactor. There will be a medium scaled-up effect when continuing to increase the flow rate. There are no flow rate monitors in this protocol, which risks an un-matched material ratio if the pump has trouble without notification.
Generally, the flow chemistry process has been shown to be superior to the batch process for synthetic transformations in the following situations: (i) using explosive or highly reactive reagents with a large safety risk (i.e., avoiding a lot of active material in the reactor); (ii) generating highly active or unstable intermediates (i.e., avoiding a long holding time before quenching); (iii) needing precise temperature control (i.e., avoiding nonuniform thermal transfer); and (iv) involving multi-liquid phase reactions (i.e., requiring enhanced mixing to accelerate the reaction rate). For those batch processes that can be transferred to the flow process, there are the benefits of increased product quality and reproducibility via enhanced and precise control of the process parameters.
Flow technology has opened new avenues for chemical pathways. Even though it may not be the universal solution to solve all the production problems that were non-viable in the batch process, from an environmental, safety and financial standpoint, it is a powerful tool to conduct multi-step processes effectively for high value-added chemicals in the pharmaceutical and fine chemical industries.
The authors have nothing to disclose.
We would like to thank the support of Shenzhen Science and Technology Program (Grant No. KQTD20190929172447117).
2-Methylpyridin-3-amine | Raffles Pharmatech Co. Ltd | C2021236-SM5-H221538-008 | HPLC: >98%, Water by KF ≤0.5% |
316L piston constant flow pump | Oushisheng (Beijing) Technology Co.,Ltd | DP-S200 | |
BF3.Et2O | Whmall.com | B802217 | |
Citric acid | Titan Technology Co., Ltd | G83162G | |
con.HCl | Foshang Xilong Huagong | 1270110101601M | |
Dynamically mixed flow reactor | Autichem Ltd | DM500 | 316L reator with 500 mL of internal volume |
Heptane | Shenzhen Huachang | HCH606 | Water by KF ≤0.5% |
Micro flow reactor | Corning Reactor Technology Co.,Ltd | G1 Galss AFR | Glass module with 9 mL of internal volume |
PTFE piston constant flow pump | Sanotac China | MPF1002C | |
Sodium hydroxide | Foshang Xilong Huagong | 1010310101700 | |
tert-Butyl methyl ether | Titan Technology Co., Ltd | 01153694 | |
tert-Butyl nitrite | Whmall.com | XS22030900060 | |
Tetrahydrofuran | Titan Technology Co., Ltd | 1152930 | Water by KF ≤0.5% |